Hello everyone,
Our IT department is pushing to set up three separate Azure tenants for Test, Production (Prod), and Pre-Prod environments. I’d like to get your thoughts on whether this is truly necessary, especially considering security, management overhead, and costs.
Security: The IT department argues that having a Public Tenant exposes our resources to the internet, which they perceive as a security risk. However, isn’t the entire concept of the Public Cloud based on the fact that resources are meant to be accessible over the internet? With the proper security configurations (such as Conditional Access, MFA, and Private Endpoints), does splitting these environments into separate tenants actually enhance security?
Wouldn't it be more efficient to implement policies at the Subscription or Management Group level to isolate environments and enforce security controls, rather than using multiple tenants, which increase management complexity?
Cost and Management: How much management overhead and extra costs are involved in maintaining multiple tenants? Are there advantages or disadvantages in terms of licensing, governance, or identity management that I should consider?
Additionally, I have a concern about the tenant’s domain name:
- If the domain name of an Azure tenant clearly reflects the company’s name (e.g.,
companyname.onmicrosoft.com
), does this create a security risk? The IT department believes that having a company-identifiable domain name increases vulnerability. However, with MFA, Conditional Access, and other security measures in place, is this really an issue? Does the domain name itself pose any significant risk if these precautions are properly implemented?
I appreciate any feedback and recommendations you may have on this. Thank you!Hello everyone,
Our IT department is pushing to set up three separate Azure tenants for Test, Production (Prod), and Pre-Prod environments. I’d like to get your thoughts on whether this is truly necessary, especially considering security, management overhead, and costs.
Security: The IT department argues that having a Public Tenant exposes our resources to the internet, which they perceive as a security risk. However, isn’t the entire concept of the Public Cloud based on the fact that resources are meant to be accessible over the internet? With the proper security configurations (such as Conditional Access, MFA, and Private Endpoints), does splitting these environments into separate tenants actually enhance security?
Wouldn't it be more efficient to implement policies at the Subscription or Management Group level to isolate environments and enforce security controls, rather than using multiple tenants, which increase management complexity?
Cost and Management: How much management overhead and extra costs are involved in maintaining multiple tenants? Are there advantages or disadvantages in terms of licensing, governance, or identity management that I should consider?
Additionally, I have a concern about the tenant’s domain name:
- If the domain name of an Azure tenant clearly reflects the company’s name (e.g.,
companyname.onmicrosoft.com
), does this create a security risk? The IT department believes that having a company-identifiable domain name increases vulnerability. However, with MFA, Conditional Access, and other security measures in place, is this really an issue? Does the domain name itself pose any significant risk if these precautions are properly implemented?
I appreciate any feedback and recommendations you may have on this. Thank you!